A U.S. federal judge recently said he turned to artificial intelligence programmes like ChatGPT to help interpret a key legal term in a man’s appeal of his more than 11 - year prison sentence. The Judge said while he was initially “spooked” by slight variances in the answers AI generated but he believed that the software could be a “valuable” tool. 

The latest case involved Joseph Deleon, who argued that a judge wrongly invoked an enhancement under the federal sentencing guidelines to his case that applies if a person was “physically restrained” to facilitate an armed robbery or an escape. 

Accused Deleon had walked into a store, pointed a gun at a cashier, demanded a register be emptied, and then left. The sentencing judge said that by doing so, Accused Deleon “physically restrained” the cashier, even though he never touched the victim. 

The Grand Jury indicted Deleon for robbery and brandishing a firearm during and in relation to a crime of violence. After a trial, a jury convicted Deleon on both count. The District Court sentenced Deleon to 240 months’ imprisonment ( 156 months on count 1 and 84 months on count 2, running consecutively) This sentence included a two-level physical-restraint enhancement. To support this enhancement, the District Court found that Deleon had physically restrained the cashier by pointing a gun at him and firing him to empty the register.  Deleon appealed the sentence but did not challenge the physical-restraint enhancement. The 11th Circuit vacated the sentence and remanded for resentencing on the ground that the District Court had added a career-offender enhancement even though Hobbs Act robbery is not a crime of violence for that purpose. On remand, the probation officer again recommended the two-level enhancement for physical restraint. Deleon again objected to the enhancement, arguing that his conduct did not rise to the level of “physical restraint”. And the District Court again rejected Deleon’s arguments and imposed the physical-restraint enhancement. That judgement Deleon appealed. 

Deleon appealed against the sentence for an armed robbery before 11th U.S. Circuit Court of Appeals, which was rejected. The Court panel consisted of Judges Robin Rosenbaum, Nancy Abudu and Kevin Newsom. The first two judges are appointees of Democratic Presidents. While the Appeal Court upheld the ruling of the sentencing judge, the judges - Rosenbaum and Abudu, in a concurring opinion said they only did so because of binding 11th Circuit Court’s precedent on the question. In those rulings, the Court had held that defendants physically restrain their victims if they create “circumstances allowing the persons no alternative but compliance”. Judge Rosenbaug said that the “full court” should take up the case to overrule those “past holdings”. Judge Rosenbaum wrote that under a “plain reading of the tex”, the cashier needed to be physically bound by Deleon, not just not psychologically as a result of the gun. Judge Newsom, an appointee of Republican President, in a separate concurring judgement agreed. Judge Newsom said because there is no dictionary definition for “physically restrained” as a combined phrase, he “couldn't help himself” and asked ChatGPT and two other generative AI programs what the phrase’s ordinary meaning is. 

Judge Kevin Newsom stated that his “humble little mini-experiment” revealed that the programs produced answers similar to an initial one ChatGPT generated, which said the phrase “refers to the act of limiting or preventing someone’s movement by using physical force or some kind of device”. Judge Newsom said he noticed slight variations in the wording and length of the answers the programs produced when asked the same question repeatedly, which initially “spooked” and “rocked” him but that he came to learn the technical explanation. Judge Newsom concluded finally that the slight variations “accurately reflects real people’s everyday speech patterns”, which he said showed the AI programs’ models could accurately predict the ordinary meaning of words. As a consequence, Judge Newsom said he increasingly believed such programs “may well serve a valuable auxiliary role as we aim to triangulate ordinary meaning”. 

Case details : United States V. Deleon, 11th Circuit Court of Appeals, No 23- 10478.

Tap Here For Judgement/ Opinion